diff -r ba454a26d2c1 -r 83e4343a6984 src/jdk.jfr/share/classes/jdk/jfr/internal/consumer/RepositoryFiles.java
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/src/jdk.jfr/share/classes/jdk/jfr/internal/consumer/RepositoryFiles.java	Thu Jun 27 10:19:32 2019 +0200
@@ -0,0 +1,126 @@
+package jdk.jfr.internal.consumer;
+
+import java.io.IOException;
+import java.nio.file.DirectoryStream;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.HashMap;
+import java.util.HashSet;
+import java.util.List;
+import java.util.Map;
+import java.util.Set;
+import java.util.SortedMap;
+import java.util.TreeMap;
+
+import jdk.jfr.internal.LogLevel;
+import jdk.jfr.internal.LogTag;
+import jdk.jfr.internal.Logger;
+import jdk.jfr.internal.Repository;
+
+public final class RepositoryFiles {
+    private final Path repostory;
+    private final SortedMap<Long, Path> pathSet = new TreeMap<>();
+    private final Map<Path, Long> pathLookup = new HashMap<>();
+    private volatile boolean closed;
+
+    public RepositoryFiles(Path repostory) {
+        this.repostory = repostory;
+    }
+
+    public long getTimestamp(Path p) {
+        return pathLookup.get(p);
+    }
+
+    public Path nextPath(long startTimeNanos) {
+        while (!closed) {
+            SortedMap<Long, Path> after = pathSet.tailMap(startTimeNanos);
+            if (!after.isEmpty()) {
+                Path path = after.get(after.firstKey());
+                Logger.log(LogTag.JFR_SYSTEM_STREAMING, LogLevel.TRACE, "Return path " + path + " for start time nanos " + startTimeNanos);
+                return path;
+            }
+            try {
+                if (updatePaths(repostory)) {
+                    continue;
+                }
+            } catch (IOException e) {
+                Logger.log(LogTag.JFR_SYSTEM_STREAMING, LogLevel.DEBUG, "IOException during repository file scan " + e.getMessage());
+                // This can happen if a chunk is being removed
+                // between the file was discovered and an instance
+                // of an EventSet was constructed. Just ignore,
+                // and retry later.
+            }
+            try {
+                synchronized (pathSet) {
+                    pathSet.wait(1000);
+                }
+            } catch (InterruptedException e) {
+                // ignore
+            }
+        }
+        return null;
+    }
+
+    private boolean updatePaths(Path repo) throws IOException {
+        if (repo == null) {
+            repo = Repository.getRepository().getRepositoryPath().toPath();
+        }
+        boolean foundNew = false;
+        List<Path> added = new ArrayList<>();
+        Set<Path> current = new HashSet<>();
+        if (!Files.exists(repo)) {
+            // Repository removed, probably due to shutdown
+            return true;
+        }
+        try (DirectoryStream<Path> dirStream = Files.newDirectoryStream(repo, "*.jfr")) {
+            for (Path p : dirStream) {
+                if (!pathLookup.containsKey(p)) {
+                    added.add(p);
+                    Logger.log(LogTag.JFR_SYSTEM_STREAMING, LogLevel.DEBUG, "New file found: " + p.toAbsolutePath());
+                }
+                current.add(p);
+            }
+        }
+        List<Path> removed = new ArrayList<>();
+        for (Path p : pathLookup.keySet()) {
+            if (!current.contains(p)) {
+                removed.add(p);
+            }
+        }
+
+        for (Path remove : removed) {
+            Long time = pathLookup.get(remove);
+            pathSet.remove(time);
+            pathLookup.remove(remove);
+        }
+        Collections.sort(added, (p1, p2) -> p1.compareTo(p2));
+        for (Path p : added) {
+            // Only add files that have a complete header
+            // as the JVM may be in progress writing the file
+            long size = Files.size(p);
+            if (size >= ChunkHeader.HEADER_SIZE) {
+                long startNanos = readStartTime(p);
+                pathSet.put(startNanos, p);
+                pathLookup.put(p, startNanos);
+                foundNew = true;
+            }
+        }
+        return foundNew;
+    }
+
+    private long readStartTime(Path p) throws IOException {
+        try (RecordingInput in = new RecordingInput(p.toFile(), 100)) {
+            ChunkHeader c = new ChunkHeader(in);
+            return c.getStartNanos();
+        }
+    }
+
+    public void close() {
+        synchronized (pathSet) {
+            this.closed = true;
+            pathSet.notify();
+        }
+    }
+}
\ No newline at end of file